A maximum likelihood approach to harmonics measurements in power systems

Abstract

Estimating the amplitudes and phases of the harmonic components of one or three-phase currents (and voltages) is required for efficient active power filtering. In this paper, we propose a new approach based on the maximum likelihood principle. This estimation is unbiased and reaches the lowest statistical variance. Moreover, its computational burden is shown to be weaker than for any other method. Two algorithms are also derived to track slowly-varying nonstationary harmonic signals.<>